[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"project-95033":3},{"id":4,"name":5,"fullName":6,"owner":7,"repo":5,"description":8,"homepage":8,"htmlUrl":8,"language":9,"languages":8,"totalLinesOfCode":8,"stars":10,"forks":11,"watchers":12,"openIssues":13,"contributorsCount":14,"subscribersCount":14,"size":14,"stars1d":14,"stars7d":15,"stars30d":15,"stars90d":14,"forks30d":14,"starsTrendScore":16,"compositeScore":17,"rankGlobal":8,"rankLanguage":8,"license":8,"archived":18,"fork":18,"defaultBranch":19,"hasWiki":20,"hasPages":18,"topics":21,"createdAt":8,"pushedAt":8,"updatedAt":22,"readmeContent":23,"aiSummary":24,"trendingCount":14,"starSnapshotCount":14,"syncStatus":25,"lastSyncTime":26,"discoverSource":27},95033,"skills","almendili\u002Fskills","almendili",null,"TypeScript",351,25,3,1,0,46,14,62.84,false,"main",true,[],"2026-08-24 04:01:23","# skills\n\nAgent skills — portable instruction packs in the `SKILL.md` format, usable by\nany harness that reads it (Claude Code, Codex, Cursor, Gemini CLI, and others).\n\n| Skill | What it does |\n|---|---|\n| [architecture-map](architecture-map\u002F) | Builds an interactive isometric map of a repository — buildings sized by real measurements, flows tracing real call paths, and a drift counter that fails CI when the map falls behind the code. |\n\n## Installing\n\nSkills are discovered from a directory your tool scans. Symlink rather than\ncopy, so `git pull` here updates every install at once:\n\n```bash\n# Claude Code (global, all projects)\nln -s \"$PWD\u002Farchitecture-map\" ~\u002F.claude\u002Fskills\u002Farchitecture-map\n\n# Codex\nln -s \"$PWD\u002Farchitecture-map\" ~\u002F.codex\u002Fskills\u002Farchitecture-map\n\n# Tools using the shared convention\nln -s \"$PWD\u002Farchitecture-map\" ~\u002F.agents\u002Fskills\u002Farchitecture-map\n\n# Scoped to one project, whatever the tool\nln -s \"$PWD\u002Farchitecture-map\" \u002Fpath\u002Fto\u002Frepo\u002F.claude\u002Fskills\u002Farchitecture-map\n```\n\nCreate the target directory first if it does not exist. Skills load when a\nsession starts, so open a new one afterwards.\n\n## What is in a skill\n\n- `SKILL.md` — YAML frontmatter (`name`, `description`) plus the instructions.\n  The description is what decides when the skill fires, so it names the\n  situations rather than describing the feature.\n- `references\u002F` — detail loaded only when needed, keeping `SKILL.md` scannable.\n- `scripts\u002F` — executables the skill runs. They live beside `SKILL.md`, not in\n  the repo being worked on, so skills resolve their own directory before\n  calling them.\n- `assets\u002F` — files copied into the target project.\n\n## Portability\n\nNothing here depends on a specific harness. Where a skill wants a structured\nquestion it names one tool's mechanism as an example and gives a plain-text\nfallback; where it needs its own directory it resolves the path rather than\nreading a tool-specific environment variable.\n","这是一个面向AI编程助手的可移植技能包框架，定义了标准化的SKILL.md格式，用于封装可复用的开发任务指令与配套脚本。核心功能包括：声明式技能元数据、按需加载的参考文档、隔离执行的本地脚本、资产自动注入，以及跨工具链（如Claude Code、Codex、Cursor、Gemini CLI）的零依赖适配能力。技术特点强调路径自解析、YAML前言驱动触发、纯文本降级兼容和符号链接式集中管理。适用于需要在多个AI编码环境间统一维护架构分析、代码审查、文档生成等自动化任务的工程团队。",2,"2026-08-20 02:30:06","CREATED_QUERY"]